Fun with vinyl measurements

[Thomas_A]

It’s correlated but a bit out of phase. Which part of the record is this?

 

[Balle Clorin]

The white noise track, track 2 side B

 

[Thomas_A]

The white noise track, track 2 side B

Ok. How close to the first null-point? Looks like zenith is a bit off.

 

[Balle Clorin]

Ok. How close to the first null-point? Looks like zenith is a bit off.

Track 2, starts at 140mm. Nullpoint should be 120.9 Never got it to match to 0 L R phase difference anywhere. Stylus maximum turned tip towards centre.( Clockwise seen from above)
Left channel arrives a bit before right, but not sure how to calculate angle error…

The expected error at 140mm would be -1 to -1.5 degree…?

Tonearm tracking error and distortion

In the last posting, I reviewed the math for calculating the tracking error for a radial tonearm. The question associated with this is “who cares?” In the March, 1945 issue of Electroni…

www.tonmeister.ca

? Ok?

 

[board]

Yes, it is a terrible thing for a stylus to be too "pointy"....

My main cartridge is an Audio Technica AT33PTG/II, but I also have an Audio Technica VM750SH, which has quite a large spike around 10-12 kHz.
On records from the 60s and 70s this often (but not always) makes a pleasant improvement, as it adds some sparkle to material that wasn't very top-end heavy to begin with.
A more modern example is Hans Zimmer's soundtrack to "The Dark Knight", which also lacked top-end.
If it's of any interest I can post audio clips to illustrate this.
But yes, it's very content specific and subjective preference as well. For music from the 80s and onwards I usually prefer a less "pointy" cartridge. Lyra, for example, I simply wouldn't buy for that reason, as well as the extortionate prices.

 

[board]

I stumbled upon some speed measurement of these very expensive DD TT. They are beaten by a typical Technics 1200…and my Gyro and my Denon 51F

Money does not always matter…

· J. · OMA K3: $363,000 including power supply and Schröder SLM tonearm · SAT XD1: €180,000 without tonearm
View attachment 385897

From what magazine is this test of the OMA and SAT?
If you have the magazine and are able to photograph and post it all (perhaps in a PM to me) I would be super grateful.
Framer seemed to make a big deal about the speed accuracy and stability of the OMA K3.

 

[Balle Clorin]

Stereophile.
Fremer makes many statements about TT that are the direct opposite of what objective measurements tell, and he is not alone. Stereophile and other magazines could easily publish Shaknspin data for any turntable tested. But I don’t think they avoid it to not embarrass the TT makers and loose income. Most TT especially high end boutique brands are really mediocre regarding speed stability. Even well know high end tables are crap if you measure them. The exceptions are SME and Reed, Technics , are a few Rega models are good if using the most expensive belt. Check out Hifi news lab data

 

[Thomas_A]

Track 2, starts at 140mm. Nullpoint should be 120.9 Never got it to match to 0 L R phase difference anywhere. Stylus maximum turned tip towards centre.( Clockwise seen from above)
Left channel arrives a bit before right, but not sure how to calculate angle error…

The expected error at 140mm would be -1 to -1.5 degree…?

Tonearm tracking error and distortion

In the last posting, I reviewed the math for calculating the tracking error for a radial tonearm. The question associated with this is “who cares?” In the March, 1945 issue of Electroni…

www.tonmeister.ca

? Ok?
View attachment 386316

I have not had time to test myself but I will. From the zoomed in wave-form you should should get silence when combined to mono if the track is well done. Thinking about it a bit more it can be related to azimuth or zenith. Same with the other table/cartridge?

 

[Balle Clorin]

Yes, Tried HP LF BP filtering too and sum and invert etc to mono , Get a weak correlation with indicication of ONE nullpoint (outer) but never a hint if the 60mm null, record stop at 70( 15 min 3150hz Clearaudio Trackability record)

As long as distortion is not the same in both channels it will not sum to null

By the way the peaks in voltage are the zero groove crossing not the groove peak. So to be correct we should compare the zero crossing to determine stylus edge phase difference L and R, but you know that…
Not sure how to take the derivative of a sinus wav file…

If I could delay one channel different usec in order to get the lowest sum /invert to mono the actual time difference may be found= time delay L and R channel= zenith error

 

[board]

Stereophile.
Fremer makes many statements about TT that are the direct opposite of what objective measurements tell, and he is not alone. Stereophile and other magazines could easily publish Shaknspin data for any turntable tested. But I don’t think they avoid it to not embarrass the TT makers and loose income. Most TT especially high end boutique brands are really mediocre regarding speed stability. Even well know high end tables are crap if you measure them. The exceptions are SME and Reed, Technics , are a few Rega models are good if using the most expensive belt. Check out Hifi news lab data

Yes, Framer (and many others) absolutely does that. Although I didn't bother to look it up now, I do remember reading Framer going on about how OMA had done blind tests of its speed stability, and differences of fractions of a second so small that it couldn't be measured made an audible difference - although I'm not necessarily convinced by yet another story from him (or from OMA), as he's been doing nothing but telling stories for 40 years. He could win the Nobel prize in literature for all the fictional stories he has come up with.
And yes, I have a Rega turntable, and although I'm happy with it I have considered buying another brand, but I've stopped looking at brands like VPI and the like. Direct drives seems more appealing to me now. Framer has actually been posting a lot of speed measurements in recent years, and I was pleasant surprised by that.
I've also been using Hifi News' website, and Reed's turntables and arms actually didn't measure as well as I had expected (but SME's did).
About Shake n' Spin, do you happen to know if it's more accurate than the test record with a 3150 Hz tone that seemed to be the standard for quite a long time?
I was under the impression that Shake n' Spin and apps like it were less accurate, but I could be mistaken.

 

[Balle Clorin]

The above is my Michell Guro SE on a good day-, it is on par with an Technics regardless type or price of price, beating TT costing 100k USD. But it need maintenance to stay that way- a DD would stay that way regardless.

I have managed to measure record WF and phone apps and overlay the results with a good match.
Shaknspin is good but not perfect. It is very good at making repeatable and consistent results. I can extract the data via Bluetooth to excel and plot it there.That is really useful for analysis.

It has 3 minor flaws.
1.To much noise so the jitter and high and low speed deviation are exaggerated . Can be solved by making a filter in excel or make a moving average.
2. 33.33 on my DD and strobe disk is reported as 33.42 on Shaknspin. It can be calibrated but that only sets whatever average speed you have to 33.33. I simply just adjust TT for 33.42.
3.The FFT plot on the display should have had a log scale in linear, it makes the FFT values to crowded to see.


4. Sometimes the DIN WF values are calculated so low that are hard to believe 0.005..

Still it is a fantastic tool to optimize and check TT rpm stability.

Most test record WF tracks are crap due to uncentered records and such. Shaknspin in calibrated shape is much better. But test records can tell more about other TT problems..


These two phone apps are the best. Thanks to them I could tune my TT from a peak wow 2S of 0.2% to 0.05…..

 

[board]

View attachment 386837I have managed to measure record WF and phone apps and overlay the results with a good match.
Shaknspin is good but not perfect. It is very good at making repeatable and consistent results. I can extract the data via Bluetooth to excel and plot it there.That is really useful for analysis.

It has 3 minor flaws.
1.To much noise so the jitter and high and low speed deviation are exaggerated . Can be solved by making a filter in excel or make a moving average.
2. 33.33 on my DD and strobe disk is reported as 33.42 on Shaknspin. It can be calibrated but that only sets whatever average speed you have to 33.33. I simply just adjust TT for 33.42.
3.The FFT plot on the display should have had a log scale in linear, it makes the FFT values to crowded to see.


4. Sometimes the DIN WF values are calculated so low that are hard to believe 0.005..

Still it is a fantastic tool to optimize and check TT rpm stability.

Most test record WF tracks are crap due to uncentered records and such. Shaknspin in calibrated shape is much better. But test records can tell more about other TT problems..


These two phone apps are the best. Thanks to them I could tune my TT from a peak wow 2S of 0.2% to 0.05…..
View attachment 386838
View attachment 386839

Do I understand you correctly that you don't have direct experience with that particular test tone record I referred to?
I think it was a 7" record made by Dr. Feickert.
I understand that test records can be problematic, but exactly this problem with Shake n' Spin showing too high a speed was the issue I thought I had heard about.
Just out of curiosity: How did you manage to change the wow on your turntable? You don't need to spend time on telling every detail - a few words is enough .
And I can imagine that this might only work on direct drives ...?

 

[Balle Clorin]

Motor not running optimal leveling

No it is a belt drive that is easy to adjust.
My Michell Gyro has a stand alone motor pad, the TT has 3 feet to adjust the height , AND spings that also affect the height, the platter has 3 grooves to select. The key point is to get the leveling and angle between motor pulley and platter such that the pull force from the belt goes trough the top of the bearing where the platter rest. That minimizes any wobbling effect.

By luck a found it the first time , but screwed it up and it took 2 years to recover.

1 . Start with a level platter ( most bubble levels are off, pick a good one where the bobble do not move when turned 180 degrees)

2 . I try all 3 platter grooves and see which has lowest WF. Do not care if leveling changes.
3. Adust feet height systematically, not more than 1/2 turn on one foot at a time .
4 write down every change.

5. When good take photos.
6. When the motor runs the belt pull changes and due to the outboard motor pulley the leveling of a gyro will change. Document that too.

7. If the TT start too lose stable WT it is time to put the belt in a bag with a tea spoon of talcum, shake and take out the belt an shake off excess talc, put the belt back on - use gloves.

I may need to relocate the motor pos a few m sometimes to compensate for belt slacking. I adjust the RPM via the motor power supply.

I hear the Feickert 7 inch wF track is quite bad….I do not have it

Same leveling motor running, notice the shift in bubble position.

For a DD you need the service manual to find the variable resistors that control the speed. For full restoration an oscilloscope is probably needed. Technics have a torque adjustment too.

EDIT: A good test record an have a wow of 0.03% a bad one 0.15-0.2 I have seen specifications of 0.06 W&F IF the record is perfectly centered

 

[board]

Thanks for your explanation. It sounds like something I wouldn't be able to do with my Rega, but it was interesting nonetheless, and I'm happy to hear that you were able to improve upon it.
As for the Dr. Feickert 7" record, I've heard stories of some people getting a really poorly pressed copy, but most copies are essentially perfectly made, as far as I understand.
So, it seems like both the apps and the test records have their benefits and downsides, as the apps are also dependent on the phones and the built-in gyroscopes, etc.

 

[undeakyetonde]

Parks Audio Waxwing dynamic range measurements, plotted together with 12, 14 and 16 bit tpdf dither noise floors

Optical output, powered from a laptop

Same cart hits around -13 to -10 dB peak RIAA output for clean 12" 33 rpm records and up to -2 dB peak for 12" 45 rpm records

Note that the noise floor in the 10k-20k range is ~15 bits

 

[Balle Clorin]

Thanks for your explanation. It sounds like something I wouldn't be able to do with my Rega, but it was interesting nonetheless, and I'm happy to hear that you were able to improve upon it.
As for the Dr. Feickert 7" record, I've heard stories of some people getting a really poorly pressed copy, but most copies are essentially perfectly made, as far as I understand.
So, it seems like both the apps and the test records have their benefits and downsides, as the apps are also dependent on the phones and the built-in gyroscopes, etc.

The Regas benefit greatly from the new super reference ELBT belts or whatever the name is. That is a must to have. you could also try the talcum trick

 

[board]

The Regas benefit greatly from the new super reference ELBT belts or whatever the name is. That is a must to have. you could also try the talcum trick

Thanks for your suggestion.
I actually bought the belt Little Belter (from a different manufacturer), although this was mainly to stabilize the speed and make the speed more accurate. I don't know if the EBLT belt would improve on that though ...
At first I had put tape around the subplatter to make it larger and thereby slow the turntable down, but in the end I peeled it off and used the Little Belter instead.
Honestly, I've been considering just replacing the whole darn turntable with a Technics 1200 or a Denon DP-3000NE, as I then don't have to worry about wow and speed stability. Making tweeks on a belt drive can seem like pouring money into a bottomless pit + it takes up a lot of time.
I am wondering if the different resonances in the Technics and Denon tonearms could cause an unpleasant tonality. Then I could mount a Rega arm instead, but then that would also take time, money and effort, so ...

 

[Balle Clorin]

I have a vintage Denon DP-51F, it is serviced with new capacitors and function perfectly. After servicing the arm the resonance at 80hz I had is reduced ,but not completely gone. I am comparing with my SME V on my Michell Gyro SE belt drive here..that arm is very good ,my measurements posted here proves it( flat and stable in the low end where many TT including modern like Technics arms show more irregularities

I the speed stability on the Denon Vintages is as good as it gets, outperforming Technics 1200G. The Denon 3000NE has not so good specs , but the sample HifiNews tested proved to be much better than the spec.

A vintage TT with better arm than the Denon is maybe the JCV QL-Y5F with and others from that family. I really like the electronic arm damping . The good thing about the Y5F is that it has a standard SME headshell and that modern headshell can be used. Very nice turntable and beating modern tables if serviced.

You can see the speed stability with the phone apps.
Any belt drive need a motor adjustment to maintain correct speed, if not you will be off most of the time, DC motors and belt will drift and change over time

 

[Balle Clorin]

Extracted the audio and compared

Not much difference

Start in groove

 

[undeakyetonde]

Parks Audio Waxwing headroom & measurements

I took:
1) a VM95ML cartridge, which is rated at 3.5 mv (1 kHz, 5 cm/sec)
2) an Ortofon Test LP, track #5 is a left-channel 5 cm/sec 1 kHz sine
3) a 45 rpm 12" single record
4) a Waxwing, of course

I recorded (24 bit / 96 kHz optical output):
1) first minute of the record (gain 38 dB)
2) the 5 cm/s sine (gain 38 dB), as another reference
3) silence with a connected cartridge (gain levels 38 dB, 28 dB, 18 dB)
*the calculator in the app recommends 40 dB gain for a 3.5 mv cartridge but I decided to stick with defaults

Silence recorded at 28 dB gain setting was digitally amplified by 10 dB
Silence recorded at 18 dB gain setting was digitally amplified by 20 dB
Sine recording is notched with a long filter (sox sine.wav sine.notched.wav sinc -n 32767 990-1010)
Everything is trimmed to exactly 96k * 60s samples.

Playing a clean 45 RPM 12" record (38 dB gain), the ADC was getting up to -3 dB peaks on transients (pre-RIAA, seen in the app). Loud pops/clicks would likely clip.

The chart below shows 20 Hz - 20 kHz smoothed spectrums (REW) for all mentioned signals, plus 16 bit tpdf dither as an additional reference
(please excuse the mains hum)

So, if we "just lower the gain", we increase headroom (of course), but receive a noise floor penalty (expected but now confirmed).

Is the penalty audible? Does it matter for real-world listening? Who knows. The background groove noise from the Ortofon sine track is about 10 dB higher [across the whole 20-20k range] than background silence of the worst gain setup (18 dB analog + 20 dB digital).